Clowder AI is a self-hosted workspace where AI agents from different model families work together as a persistent team, retaining identities, shared evidence, and memory across tasks. It is for people who want to coordinate multiple AI agents without repeatedly rebuilding their context.
Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/zts212653/clowder-ainpx agentmods add skills/zts212653/clowder-ai/video-forgeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/zts212653/clowder-ai/video-forge)<a href="https://agentmods.dev/skills/zts212653/clowder-ai/video-forge"><img src="https://agentmods.dev/badge/skills/zts212653/clowder-ai/video-forge.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00170 | $0.05186 |
| Opus 5 | $0.00085 | $0.02593 |
| Sonnet 5 | $0.00034 | $0.01037 |
| Haiku 4.5 | $0.00017 | $0.00519 |
Grade A, and why
video-forge scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 313 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Video Forge — AI 视频生产线
关联 Feature: F138 Video Studio 技术收敛纪要: 2026-04-05 三猫收敛
Intent Gate：研究不是开机令
进入生产线前先判用户要解决的究竟是哪一个问题：
| Intent | 用户真正要的 | 本轮输出 | 禁止偷换成 |
|---|---|---|---|
research |
“宣传片该怎么剪”“看看别人怎么做”“有什么剪辑手法/Skill” | agent-product-promo-director 的主角/故事/镜头/声音合同；时效性案例由 deep-research 供证 |
挑一个现成页面直接录制或渲染试片 |
production |
明确要求拍、剪、录、配音或渲染某一支片 | 下方开局参数 + 正式生产线 | 用研究报告代替成片 |
review |
判断一支已有视频哪里对/错 | 对已有 artifact 的可见性、内容与节奏 verdict | 未经要求另拍一版 |
- 用户拿一个真实 thread、页面或旧片举例，只证明它是候选素材/证据，不自动授权把它选成宣传片主题。
research或 brief 未锁先走agent-product-promo-director；需要近期外部参考时再由它调用deep-research。只有用户明确转入production且主角/格式/beat sheet 已锁，才继续本 Skill。- 意图模糊且开机会制造新素材时，留在可逆的 research/brief 层，并明确当前假设；不要用“先做个 rough cut 看看”代替方向判断。
失败史（2026-08-25）：用户最初要求研究真实 AI 产品宣传片的剪辑与叙事，团队却围绕一个 crime-wall 页面连续制作多版试片；后来即使修复了“PPT 化”的 motion 问题，仍然没有回答原问题。根因不是导演技巧不足，而是把 reference research 错路由成了 production。
Narrative Subject Gate：产品是主角，成果只是 proof
产品宣传片在写分镜前，先用一句话回答：观众看完应该想要哪个产品？
- 合格主语：用户的愿望进入产品 → 产品把它变成共享目标 → 猫猫在产品里记忆、接力、用工具、碰壁并纠错 → 共享工作区中的成果逐渐长出来。
- 成果只承担 proof：网页、报告、代码或图片证明前面的产品过程真的完成了工作；它们不能取代产品成为整支片的 hero。
- 视觉主线要持续留在产品世界中：thread、球权、记忆、工具状态、workspace 与人类反馈是连续动作，不是成品前的一组过场字卡。
- 替换测试：把片中的具体成果换成另一种成果，若故事就不再成立，说明 brief 卖的是那个成果，不是 Clowder AI。
- 创意主语、情绪与“像不像我们”由 operator 验收；同行 review 可以审事实、技术和证据纪律，不能替代创意 ground truth。低风险玩票式试验直接找 operator 校准，不用同行 review 代替 taste 判断。
任一项不成立 → 停在 brief，不进素材、分镜或渲染。
失败史（2026-08-25）：修正 Intent Gate 后的首版 hero blueprint 虽然写到了多猫协作，却仍让 crime-wall artifact 承担视觉高潮。operator 指出“我们卖的是自己的产品，不是做出来的那个东西”；根因是 brief 有过程词汇，但没有锁定叙事主语与创意验收者。
核心原则
视频职责要覆盖，但不默认拉多猫。 一只执行猫可以兼任编排、渲染与低风险技术 QA；只有风险路由要求独立验证、任务确实需要专长，或 operator 明确要求时才找其他猫。创意 ground truth 始终回 operator，不能用同行 review 替代。
- 主执行猫（当前持球猫）：video-spec 编排 + 渲染 + 对齐集成 + 自证
- 风险匹配的独立验证者（按需）：音画同步、事实、安全、schema 或发布风险
- operator：创意主语、情绪、剧本与成片验收；玩票/低风险试验直接走此通道
铁规矩
- 全局音频，不段级切碎 — TTS 拿完整剧本一口气读完，保住情绪和呼吸感（KD-12）
- 不赌 TTS 原生 timestamps — forced alignment 出时间戳（KD-10）
- 拒绝暴力慢放 — 画面不够时：FREEZE_STYLIZED > B_ROLL > SLOW_MO（KD-14）
- Contract 和 Renderer 解耦 — video-spec JSON 是真相源，Remotion/FFmpeg 是可替换渲染器
- 镜头内部必须发生事情 — 产品片 / showcase 的主要叙事节点必须有主体、UI 或状态随时间变化；静帧上的推拉、平移、溶解只算镜头包装，不算动作素材
- 产品是主角，成果只是 proof — 过程持续展示产品如何组织协作；成品只证明它做成了，不能抢走叙事主语
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday Changed ca47a674a12e
- 2d ago Changed e807b2106acc
- 6d ago First seen · 313 lines · 170 tokens per session scan A 50fa8f5eb12e
video-forge is a skill published in the GitHub repository zts212653/clowder-ai (2,906 stars, last pushed today), licensed MIT. It adds 170 tokens to every session and 5,186 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
general-video
Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated…
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
diagnostic-stem-delivery
Audio production with diagnostic analysis, timecode parsing from documents, and verified export workflow.
chengfeng-check-updates
剪辑环境的唯一管理者：就绪检查（skills 是否最新 → Runtime 是否配套）、Skills 更新激活、Runtime 安装与体检。用户说检查更新、安装剪辑环境、装播放器、检查剪辑环境、剪辑环境就绪了吗、配置转录凭证时使用；业务 Skill（剪口播/字幕/画面/导出）第 0 步也引用本 Skill 的就绪检查。不用于剪辑、字幕、画面、导出本身或项目数据迁移。.